As the development of media applications, more and more applications for editing audio data appear, and are no longer limited to implementing collection and playing of various pieces of audio data, but are more focused on edition of the audio data. A variety of audio data outputs are implemented by using collected audio data as an original sound.
For example, in a karaoke function provided for a user, the collected audio data are mostly edited and attached with a certain sound effect before outputting for a listener to listen. During edition of the audio data, if simulation of a harmonic sound effect is required to be implemented, the corresponding implementing mode is to use the collected audio data, i.e., the input human voice as the original sound, and superpose other human voices at a certain number and a certain ratio onto the original sound on this basis so as to obtain audio data to be output.
However, the implementation of the harmonic sound effect simulation is to attach the harmonic sound effect to the entire original sound, i.e., adding harmonic during the entire time period, which is contradict with the harmonic in a real scene, and has a limitation that the harmonic sound effect simulation is distorted.